Origin
I was catching up with a friend, who is also a colleague, for the
first time in quite awhile in early March.
We were texting over whatsapp and we were discussing the difficulties
that she has using our video conferencing service now that all her
meetings are remote as she is getting increasingly hard of hearing.
We went on to discuss how some of the apps do real-time
transcription/close-captioning. However, the one the company uses,
primarily, doesn't. I then had the thought how easy it should be too:
What I built
- create a web page where the user could enter the audio dial in for their call
- have Twilio dial that number
- pass the audio stream to Google Speech to Text
- feed closed captions back to the user
Category Submission:
COVID-19 Communications: an app to help folks communicate during this challenging time, either to stay connected or to help resolve COVID-19 related challenges.
Demo Link
Link to Code
How I built it (what's the stack? did I run into issues or discover something new along the way?)
The application is built on nodejs and express. The entire application runs in a container for easy deployment and scalability.
I think the biggest problems I ran in to were those typical "gaps in documentation." The items are probably documented somewhere but it took a long time to find because I was trying to stand something up that was new.
Additional Resources/Info
While doing this, I need the Twilio CLI and I hate installing things...
If you need the Twilio CLI but don't want to actually install it, I also wrote a "twilio-helper" container that should be handy.
Acknowledgements
The application is largely based on work from nokenwa's demo.
As always, many, many references on the internet.
I think I have about 70 odd tabs open.
Top comments (1)
I didn't realize at first that "demo" could be a video, I was thinking it had to be live. Well, I made a video demo to make it clearer how the applications works. Check it out!